Fixed-flexion knee radiography: the sensitivity to detect knee joint space narrowing in osteoarthritis

Abstract

Independent reports on the longitudinal performance of non-fluoroscopic fixed-flexion knee radiography have not been published. Therefore, we evaluated the sensitivity of fixed-flexion radiography to detect knee joint space narrowing (JSN) over 2 years in patients with osteoarthritis (OA) and compared the effect of reproducibility and quality of medial tibial plateau (MTP) alignment on sensitivity to JSN. Fixed-flexion radiographs of both knees of 193 OA patients were obtained at baseline and after 24 months. Minimum joint space width (JSW) of the medial tibiofemoral joint spaces was measured manually in paired digitised radiographs without knowledge of the chronology. The degree of MTP alignment was assessed by measuring the distance between the anterior and posterior margins of the MTP (intermargin distance [IMD]). Accurate repositioning was achieved if, in the 24-month radiograph, the IMD was reproduced to within 1mm. The quality of MTP alignment was satisfactory if the IMD was <or=1mm. The mean (standard deviation [SD]) JSN between baseline and 24 months was 0.18 (0.49)mm and the standardised response mean (SRM) was 0.37. The SRM was 0.36 in the knee film pairs (86%) in which the IMD was accurately reproduced after 24 months and 0.52 in the film pairs (42%) with satisfactory MTP alignment on both the baseline and 24-month radiographs. The sensitivity to radiographic JSN on fixed-flexion knee radiographs over 2 years was higher in satisfactorily aligned than accurately reproduced serial pairs of radiographs. These findings have implications for longitudinal knee OA studies using fixed-flexion radiographs.